Virtual Intership KPMG

KPMG’s Analytics, Information & Modelling group helps organisations take the mystery out of big data and show them how to leverage their data resources to produce better business outcomes.

Task 2

Data Insights
Data is from Sprocket Central Pty Ltd , a medium size bikes & cycling accessories organisation.

Using the existing 3 datasets (Customer demographic, customer address and transactions) as a labelled dataset, please recommend which of these 1000 new customers should be targeted to drive the most value for the organisation.

Libraries

Load the Data

We have cleaned the data in the before module.

Final Data

We shall combine all the given data to one single datasheet for final analysis

Final Data

Let's merge the meta data(CustomerDemgraphic) and transactions data for the final analysis

We'll have a quite a few CustomerID missing coz the transaction data does not have all the ID's.

Clustering

Lets cluster our customers on the basis of the number of products bought(frequency), the total number of profit earned by the company from the customer and thier recent purchase date.

Value-based segmentation

[(Profit (frequency / 100)) / latest_purchase_day]*

Analysis

Key Points

  1. Their is a equal distribution between men and woman.
  2. High quantity of customers present in New South Whales
  3. 40-50 age category seems to be higher than others.
  4. Our main three customers are from health, manufacturing and financial services industries

There is equal distribution of men and woman in industries and New South Wales

Greater Number of frequency more the profit
More Frequencies are around 4-6

In all three ranks Solex and WeareA2B brand is bought the most

Conclusion

  1. As per the clusters the company can focus on the Royal ,Loyal and Friendly rank in future for increase in profit.
  2. Solex brand is bought the most.
  3. Most customers come from Health, Finance and Manufacturing industries.